{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/illuminating-pedestrians-via-simultaneous","title":"Illuminating Pedestrians via Simultaneous Detection & Segmentation","arxiv_id":"1706.08564","date":"2017-06-26","proceeding":"ICCV 2017 10","authors":["Garrick Brazil","Xi Yin","Xiaoming Liu"],"abstract":"Pedestrian detection is a critical problem in computer vision with\nsignificant impact on safety in urban autonomous driving. In this work, we\nexplore how semantic segmentation can be used to boost pedestrian detection\naccuracy while having little to no impact on network efficiency. We propose a\nsegmentation infusion network to enable joint supervision on semantic\nsegmentation and pedestrian detection. When placed properly, the additional\nsupervision helps guide features in shared layers to become more sophisticated\nand helpful for the downstream pedestrian detector. Using this approach, we\nfind weakly annotated boxes to be sufficient for considerable performance\ngains. We provide an in-depth analysis to demonstrate how shared layers are\nshaped by the segmentation supervision. In doing so, we show that the resulting\nfeature maps become more semantically meaningful and robust to shape and\nocclusion. Overall, our simultaneous detection and segmentation framework\nachieves a considerable gain over the state-of-the-art on the Caltech\npedestrian dataset, competitive performance on KITTI, and executes 2x faster\nthan competitive methods.","url_abs":"http://arxiv.org/abs/1706.08564v1","url_pdf":"http://arxiv.org/pdf/1706.08564v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"illuminating-pedestrians-via-simultaneous","repo_url":"https://github.com/Ricardozzf/sdsrcnn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"illuminating-pedestrians-via-simultaneous","repo_url":"https://github.com/garrickbrazil/SDS-RCNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"autonomous-driving","task_name":"Autonomous Driving"},{"task_slug":"pedestrian-detection","task_name":"Pedestrian Detection"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/pedestrian-detection-on-caltech","task":"Pedestrian Detection","dataset":"Caltech","model":"SDS-RCNN","rank_in_archive_order":20,"of":33,"metrics":{"Reasonable Miss Rate":"7.36"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.08564","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}